Susceptibility-Weighted Imaging of the Anatomic Variation of Thalamostriate Vein and Its Tributaries

Background and Purpose Thalamostriate vein (TSV) is an important tributary of the internal cerebral vein, which mainly drains the basal ganglia and deep medulla. The purpose of this study was to explore the anatomic variation and quality of TSV and its smaller tributaries using susceptibility-weighted imaging (SWI). Methods We acquired SWI images in 40 volunteers on a 3.0T MR system using an 8-channel high-resolution phased array coil. The frequencies of the TSV and its tributaries were evaluated. We classified TSV into types I (forming a venous angle) and II (forming a false venous angle). We classified anterior caudate vein (ACV)into types 1 (1 trunk) and 2 (2 trunks) as well as into types A (joiningTSV), B (joining anterior septal vein), and C (joining the angle of both veins). Results The TSV drains the areas of caudate nucleus, internal capsule,lentiform nucleus, external capsule, claustrum, extreme capsule and the white matter of the frontoparietal lobes,except thalamus. The frequencies of the TSV, ACV and transverse caudate vein (ACV) were 92.5%, 87.5% and 63.8%, respectively. We found TSV types I and II in 79.7%, and 20.3% with significantly different constitution ratios (P< 0.05). The most common types of ACV were type 1 (90.0%) and type A (64.3%). Conclusion The complex three-dimensional (3D) venous architecture of TSV and its small tributaries manifests great variation, with significant and practical implications for neurosurgery.


Introduction
Susceptibility-weighted imaging (SWI) is a relatively new MR imaging technique based on variation in blood oxygenation between venous blood andsurrounding cerebral parenchyma [1]. The differences in magnetic susceptibility between oxygenated and deoxygenated haemoglobin are delineated by SWI in terms of large (diameter approximately 1mm) and small (diameter less than 1 mm) veins in the brain using a long TE, 3D gradient-echo MR sequence [2]. Digital subtraction angiography (DSA) remains the gold standard for measurement of vessel dimensions invivo, but is an invasive technique showing unilateral veins [3]. The most widely used and relatively appropriate technique is magnetic resonance venography (MRV) which is limited in its ability to visualize small vessels [4]. SWI is significantly superior to MRV [4] with regard to smaller venous structures and DSA [5], without the need for intravenous contrast agent. In recent decades, several studies have used the technique to investigate deep cerebral veins [5,6], cerebellar veins [7]and spinal veins [8,9].
Thalamostriate vein (TSV) is the largest tributary of internal cerebral vein, which mainly drains the areas of basal nuclei and frontoparietal white matter. In neurosurgery of the third ventricle, TSV or theforamen of Monro is frequently used as a ventricular landmark [10,11]. We often occluded TSV for a wider exposure to the third ventricle [12]. However, Mohamed et al. [13] suggest that deliberate occlusion of TSV should be performed only when absolutely necessary to avoid the risk of infarcts of the basal nuclei. Therefore, a thorough understanding of the anatomic variation of TSV and its tributaries is imperative.
To our knowledge, studies investigating the anatomy of TSV and its tributaries are limited. The purpose of this study was to explore the normal anatomy of TSV and its anatomic variations and small tributaries using SWI invivo.

Volunteer Selection
A total of 40 healthy adult volunteers (22 Females and 18 males; age range, 20-35; mean age, 26) were included. The absence of cerebral and other intracranial diseaseswas confirmed in all volunteers. All participants signed informed consent and were informed of the potential side effects of 3.0T MRI, including vertigo, nausea, and claustrophobia. The study design was approved by the Ethics Committee of Wenzhou Medical University.

MRI
All healthy volunteers were examined on a 3T MR system (Royal Philips Electronics, Amsterdam, The Netherlands)using an 8-channelhigh-resolution phased array coil. The following sequences were performed: (1)T1-weighted imaging(T1WI) and fluid-attenuated inversion recovery(FLAIR) sequence

Image Processing
Images were processed using the Extended MR WorkSpace release 2.6.3.4 workstation (Philips). The SWI images were reformatted using minimum intensity projections (mIPs) technique with contiguous 20 mm-thick sections and -19 mm section gap in the transverse, sagittal and coronal planes. The final thickness of the sections was 1 mm.

Evaluation
The bilateral TSV and small tributaries were evaluated bythree experienced neuroradiologistsin the transverse, sagittal and coronal planes of SWI. We classified the anatomic variation of TSV and anterior caudate veins (ACV) as follows.

Evaluation of TSV
We classified TSV into 2 types called Type I and Type II based on a previous study [14].
Type I: The U-shaped junction of TSV and the internal cerebral vein was adjacent to the posterior margin of the foramen of Monro forming the venous angle (Fig 1).
Type II: The U-shaped junction of TSV and the internal cerebral vein extended beyond the posterior margin of the foramen of Monro forming a false venous angle (Fig 1).  (Fig 3). Type A involvesACVs joining TSV. In type B, the ACVs join the anterior septal vein (ASV). The type C ACVs are attached to the angle of TSV and ASV.

Statistical Analysis
The differences in anatomic variation of both hemispheres were assessed using the χ 2 test, with a value of P< 0.05 considered statistically significant. Data were analyzed using SPSS version 16.0 (IBMSPSS, Chicago, IL, USA).

Results
Using mIP processing software and multiplanar reconstruction (MPR), SWI clearly showed TSV and its small tributaries (Fig 4a and 4b). The detection rates of each vein are summarized in Table 1. The TSV originates in the anterior wall of the atrium or in the inferolateral wall of the body of the lateral ventricle, and passing anteriorly and medially, beneath the stria terminalis it receives smaller tributaries on its way toward the foramen of Monro. The drainage areas of TSV include the caudate nucleus, internal capsule, lentiform nucleus, external capsule, claustrum, extreme capsule and the white matter of the frontoparietal lobes (Fig 4a and 4b). There were few veins of thalamus draining into TSV. Inthe transverse view, we also found small deep medullary veinswhich anastomose the tributaries of thalamostriate vein with the superficial medullary veins (Fig 4c and 4d).  Table 2 summarizes the frequencies of the two types. In 59(79.7%) of 74 sides, the venous angle was formed (Fig 5a; Table 2). In 15(20.3%) of 74 sides, the false venous angle was formed (Fig 5b; Table 2). There were significant differences in the constitution ratios between the right and left hemispheres (P = 0.018 <0.05).

Discussion
SWI is a novel neuroimaging technique, based on differences in tissue magnetic susceptibility yielding a unique contrast, different from that of spin density, T1, T2, and T2 Ã [15]. It displays small anatomic structures such as veins and deep nuclei with superb contrast and resolution, based on increased iron content, unlike conventional MRI. Compared with other imaging techniques, SWI does not require injection of contrast material. It is sensitive to veins, especially small veins. SWI is free from arterial contamination which is a major challenge with MRV. Rapid advances in medical imaging have resulted in the emergence ofultrahigh-field MRI systems of 7.0T [16][17][18]and 9.0 T [19]. SWI reveals deep cerebral venous anatomy [6,20]. Itis a reliable tool to measure the cerebral venous diameter [5].  SWI is comparable to 3D CE-MRV in revealing TSV anatomy and provides better image resolution of ACVs [21]. In our study, the TSV and all the smaller tributaries were clearly visualized. We found the TSV, ACV and TCV in 92.5%, 87.5% and 63.8% of all hemispheres, respectively. Michio Ono [22]found the TSV and ACV in 90% and 100% of all autopsy samples, respectively. Compared with the anatomic study, the frequency of TSVwas high, but the ACV detection rate was low in our study, probably due to sample differences. The SW images showed that the TSV received small veins from the caudate nucleus, internal capsule, lentiform nucleus and the deep white matter of the frontoparietal lobes, but not from the thalamus, consistent with previousstudies [6,23,24]. Therefore, the TSV may be a misnomer. TSV might bebetter designated as frontoparietal internal vein,instead, as proposed byHassler [23].
In addition,we know that the abundance of TSV collateral circulation contributes to its safe occlusion in neurosurgery. Several deep medullary veins, which anastomose with the superficial medullary veins drain into the TSV. The small size of the anastomotic veins prevents detection with routine vascular imaging techniques. Juliane Budde [25]reported that these small veins were clearly demonstrated at 9.0T rather than 3.0 T. In our study, numerous small transmedullary veins were clearly visible in the high-resolution images from 3.0 T (Fig 4d). The improved detection of tiny veins in our study may be related to better parameter setting and longer scanning time.
The junction of the TSV and internal cerebral vein (ICV)was seen adjacent to the posterior margin of the foramen of Monro called the venous angle or beyond it called the false venous  angle. The foramen of Monro beyond the TSV-ICV junction can be enlarged to provide a wider access to the third ventricle without occluding the TSV. Several studies have described varying frequencies of the venous or false venous angle. Shinya Fujii [6]found the false venous angle in 19.1% using phase-sensitive imaging whereas N. Cagatay Cimsit [26] detected it in 34% of all subjects using MR time of flight(TOF) venography. The frequency of a false venous angle was 32.5% [14] or 39% [27] in two anatomical dissection studies. We observed the false venous angle in 20.3%. Therefore, the venous angle is the most common type suggesting possible damage to TSV during the neurosurgery of third ventricle. We also found a statistically significant difference in the ratios of venous angle in both sides (P<0.05), more commonly in the right hemispheres than the left. It suggests that abnormally enlarged TSV in the right side may pose a higher risk of obstruction of the foramen of Monro. Although few reports suggest vascular lesions as the cause of obstruction of the foramen of Monro, Jody Leonardo [28] reported a rare case report, which displayed an abnormally enlarged TSV, leading to a unilateral obstruction and right-sided hydrocephalus. The U junction of the false venous angle was  found to be 4.3mm (3-7mm) beyond the posterior margin of the foramen of Monro in a previous anatomical study [14] whereas it was9.15 ± 4.09mm (3.20-14.00mm)in our study. The ACVs originate in the ventricular aspect of the head of the caudate nucleus and always converge posteriorly to form 1 or 2 trunks to join the TSV stem [6]. In our study, most of the ACVs (90.0%) formed a single trunk and 64.3% of all ACVs joined the stem of TSV. The frequency of types B or C, which join the ASV or the angle between ASV and TSV was low compared with type A, without any significant differences in the ratios bilaterally.

Conclusion
The venous architecture of TSV and its small tributaries is a complex three-dimensional (3D) network manifesting great variation. A thorough preoperative understanding of the variation offers significant practical neurosurgical guidance.